This is based on one of the fixes from Igor's series: 20250207162048.1890669-1-imamm...@redhat.com
Rather than reverting the patch I cleaned up a couple of cases where we either didn't need to do a flush or we could trivially defer to when the CPU started running. For the other cases I introduced tlb_flush_other_cpu() for users like reset and post migration. Any other users wanting to do a cross-vcpu blat everything function like tlb_flush() are almost certainly wanting to use a different more targeted helper which still supports queuing async work. Alex. Alex Bennée (3): target/ppc: drop ppc_tlb_invalidate_all from cpu_reset target/hppa: defer hppa_ptlbe until CPU starts running cputlb: introduce tlb_flush_other_cpu for reset use Igor Mammedov (1): tcg:tlb: use tcg_debug_assert() in assert_cpu_is_self() include/exec/exec-all.h | 20 ++++++++++++++++---- accel/tcg/cputlb.c | 18 +++++++++++++----- accel/tcg/tcg-accel-ops.c | 2 +- cpu-target.c | 2 +- target/hppa/cpu.c | 10 +++++++++- target/i386/machine.c | 2 +- target/ppc/cpu_init.c | 3 --- 7 files changed, 41 insertions(+), 16 deletions(-) -- 2.39.5
